\begin{table}[H] \centering
\newcolumntype{R}{>{\raggedleft\arraybackslash}X}
\newcolumntype{L}{>{\raggedright\arraybackslash}X}
\newcolumntype{C}{>{\centering\arraybackslash}X}

\caption{Effects of electoral turnovers taking place in adverse economic conditions}
\label{tab:baseline_het_oil_shock}
\begin{tabularx}{\linewidth}{@{}lCCCCC@{}}

\toprule
 & &  \multicolumn{3}{c}{\textbf{Adverse oil price growth}} & \textbf{Global shocks} \tabularnewline & (1) & (2) & (3) & (4) & (5) \tabularnewline 
{}&{Baseline}&{Above median}&{Top tercile}&{Top quartile}&{Post 73/79/07} \tabularnewline
\midrule \addlinespace[\belowrulesep]
\textbf{Economic performance}&0.269***&0.169&0.296**&0.244*&0.436*** \tabularnewline
&(0.101)&(0.132)&(0.160)&(0.146)&(0.188) \tabularnewline
\textbf{HDI}&0.197&0.174&0.190&-0.055&0.255 \tabularnewline
&(0.167)&(0.262)&(0.330)&(0.373)&(0.494) \tabularnewline
\textbf{Democracy}&0.192**&0.039&0.000&-0.064&-0.017 \tabularnewline
&(0.101)&(0.175)&(0.189)&(0.219)&(0.195) \tabularnewline
\textbf{General index}&0.276***&0.146&0.190&0.127&0.394* \tabularnewline
&(0.105)&(0.131)&(0.164)&(0.164)&(0.223) \tabularnewline
\bottomrule \addlinespace[\belowrulesep]

\end{tabularx}
\\ \parbox{\linewidth}{\footnotesize \caption*{\footnotesize \emph{Notes}: This table reports estimated effects of electoral turnovers for the full sample in column (1), and for the subsamples of elections with adverse oil price growth above the 50th, 66.6th, and 75th percentiles (in columns 2 to 4). These percentiles are measured among elections with a running variable between -15pp and +15pp. The adverse oil price growth variable is the average yearly growth in oil prices in the two years before the election for net oil importers, and the opposite of this variable for net oil exporters (see Section \ref{sec:externalValidity} for details). In column (5), we restrict the sample to elections taking place in the four years following three global economic shocks: the 1973 and 1979 oil crises in oil-importing countries, and the 2007-08 financial crisis. Namely, we look at elections taking place between 1974-77 and 1980-83 in net oil-importing countries, as well as in 2008-11, worldwide. $^{*} p<0.10,^{**} p<0.05,^{***} p<0.01$.}}
\end{table}
